ViewVC Help

View File | Revision Log | Show Annotations | Root Listing

root/cvsroot/COMP/CRAB/python/cms_cmssw.py

Comparing COMP/CRAB/python/cms_cmssw.py (file contents):
Revision 1.8 by gutsche, Wed Jun 14 20:52:32 2006 UTC vs.
Revision 1.116.2.2 by fanzago, Tue Oct 9 17:18:11 2007 UTC

#	Line 2 | Line 2 | from JobType import JobType
2		from crab_logger import Logger
3		from crab_exceptions import *
4		from crab_util import *
5	+	from BlackWhiteListParser import BlackWhiteListParser
6		import common
6	–	import PsetManipulator
7	–
8	–	import DBSInfo_EDM
9	–	import DataDiscovery_EDM
10	–	import DataLocation_EDM
7		import Scram
8
9	<	import os, string, re
9	>	import os, string, glob
10
11		class Cmssw(JobType):
12	<	def __init__(self, cfg_params):
12	>	def __init__(self, cfg_params, ncjobs):
13		JobType.__init__(self, 'CMSSW')
14		common.logger.debug(3,'CMSSW::__init__')
15
20	–	self.analisys_common_info = {}
21	–	# Marco.
16		self._params = {}
17		self.cfg_params = cfg_params
18	+
19	+	# init BlackWhiteListParser
20	+	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
21	+
22	+	try:
23	+	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
24	+	except KeyError:
25	+	self.MaxTarBallSize = 9.5
26	+
27	+	# number of jobs requested to be created, limit obj splitting
28	+	self.ncjobs = ncjobs
29	+
30		log = common.logger
31
32		self.scram = Scram.Scram(cfg_params)
27	–	scramArea = ''
33		self.additional_inbox_files = []
34		self.scriptExe = ''
35		self.executable = ''
36	+	self.executable_arch = self.scram.getArch()
37		self.tgz_name = 'default.tgz'
38	+	self.additional_tgz_name = 'additional.tgz'
39	+	self.scriptName = 'CMSSW.sh'
40	+	self.pset = ''      #scrip use case Da
41	+	self.datasetPath = '' #scrip use case Da
42
43	+	# set FJR file name
44	+	self.fjrFileName = 'crab_fjr.xml'
45
46		self.version = self.scram.getSWVersion()
47	+
48	+	#
49	+	# Try to block creation in case of arch/version mismatch
50	+	#
51	+
52	+	a = string.split(self.version, "_")
53	+
54	+	if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
55	+	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
56	+	raise CrabException(msg)
57	+	if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
58	+	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
59	+	raise CrabException(msg)
60	+
61	+	common.taskDB.setDict('codeVersion',self.version)
62		self.setParam_('application', self.version)
36	–	common.analisys_common_info['sw_version'] = self.version
37	–	### FEDE
38	–	common.analisys_common_info['copy_input_data'] = 0
39	–	common.analisys_common_info['events_management'] = 1
63
64		### collect Data cards
65	+
66	+	## get DBS mode
67		try:
68	<	#   self.owner = cfg_params['CMSSW.owner']
69	<	#   log.debug(6, "CMSSW::CMSSW(): owner = "+self.owner)
70	<	#   self.dataset = cfg_params['CMSSW.dataset']
71	<	self.datasetPath = cfg_params['CMSSW.datasetpath']
72	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+self.datasetPath)
68	>	self.use_dbs_1 = int(self.cfg_params['CMSSW.use_dbs_1'])
69	>	except KeyError:
70	>	self.use_dbs_1 = 0
71	>
72	>	try:
73	>	tmp =  cfg_params['CMSSW.datasetpath']
74	>	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
75	>	if string.lower(tmp)=='none':
76	>	self.datasetPath = None
77	>	self.selectNoInput = 1
78	>	else:
79	>	self.datasetPath = tmp
80	>	self.selectNoInput = 0
81		except KeyError:
49	–	#    msg = "Error: owner and/or dataset not defined "
82		msg = "Error: datasetpath not defined "
83		raise CrabException(msg)
84
85		# ML monitoring
86		# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
87	<	datasetpath_split = self.datasetPath.split("/")
88	<	self.setParam_('dataset', datasetpath_split[1])
89	<	self.setParam_('owner', datasetpath_split[-1])
87	>	if not self.datasetPath:
88	>	self.setParam_('dataset', 'None')
89	>	self.setParam_('owner', 'None')
90	>	else:
91	>	try:
92	>	datasetpath_split = self.datasetPath.split("/")
93	>	# standard style
94	>	self.setParam_('datasetFull', self.datasetPath)
95	>	if self.use_dbs_1 == 1 :
96	>	self.setParam_('dataset', datasetpath_split[1])
97	>	self.setParam_('owner', datasetpath_split[-1])
98	>	else:
99	>	self.setParam_('dataset', datasetpath_split[1])
100	>	self.setParam_('owner', datasetpath_split[2])
101	>	except:
102	>	self.setParam_('dataset', self.datasetPath)
103	>	self.setParam_('owner', self.datasetPath)
104	>
105		self.setTaskid_()
106		self.setParam_('taskId', self.cfg_params['taskId'])
107
61	–
62	–
63	–
108		self.dataTiers = []
65	–	#       try:
66	–	#           tmpDataTiers = string.split(cfg_params['CMSSW.data_tier'],',')
67	–	#           for tmp in tmpDataTiers:
68	–	#               tmp=string.strip(tmp)
69	–	#               self.dataTiers.append(tmp)
70	–	#               pass
71	–	#           pass
72	–	#       except KeyError:
73	–	#           pass
74	–	#       log.debug(6, "Cmssw::Cmssw(): dataTiers = "+str(self.dataTiers))
109
110		## now the application
111		try:
#	Line 90 | Line 124 | class Cmssw(JobType):
124		try:
125		self.pset = cfg_params['CMSSW.pset']
126		log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
127	<	if (not os.path.exists(self.pset)):
128	<	raise CrabException("User defined PSet file "+self.pset+" does not exist")
127	>	if self.pset.lower() != 'none' :
128	>	if (not os.path.exists(self.pset)):
129	>	raise CrabException("User defined PSet file "+self.pset+" does not exist")
130	>	else:
131	>	self.pset = None
132		except KeyError:
133		raise CrabException("PSet file missing. Cannot run cmsRun ")
134
135		# output files
136	+	## stuff which must be returned always via sandbox
137	+	self.output_file_sandbox = []
138	+
139	+	# add fjr report by default via sandbox
140	+	self.output_file_sandbox.append(self.fjrFileName)
141	+
142	+	# other output files to be returned via sandbox or copied to SE
143		try:
144		self.output_file = []
101	–
145		tmp = cfg_params['CMSSW.output_file']
146		if tmp != '':
147		tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
#	Line 108 | Line 151 | class Cmssw(JobType):
151		self.output_file.append(tmp)
152		pass
153		else:
154	<	log.message("No output file defined: only stdout/err will be available")
154	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
155		pass
156		pass
157		except KeyError:
158	<	log.message("No output file defined: only stdout/err will be available")
158	>	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
159		pass
160
161		# script_exe file as additional file in inputSandbox
162		try:
163	<	self.scriptExe = cfg_params['USER.script_exe']
164	<	self.additional_inbox_files.append(self.scriptExe)
163	>	self.scriptExe = cfg_params['USER.script_exe']
164	>	if self.scriptExe != '':
165	>	if not os.path.isfile(self.scriptExe):
166	>	msg ="ERROR. file "+self.scriptExe+" not found"
167	>	raise CrabException(msg)
168	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
169		except KeyError:
170	<	pass
171	<	if self.scriptExe != '':
172	<	if os.path.isfile(self.scriptExe):
173	<	pass
174	<	else:
175	<	log.message("WARNING. file "+self.scriptExe+" not found")
176	<	sys.exit()
130	<
170	>	self.scriptExe = ''
171	>
172	>	#CarlosDaniele
173	>	if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
174	>	msg ="Error. script_exe  not defined"
175	>	raise CrabException(msg)
176	>
177		## additional input files
178		try:
179	<	tmpAddFiles = string.split(cfg_params['CMSSW.additional_input_files'],',')
179	>	tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
180		for tmp in tmpAddFiles:
181	<	if not os.path.exists(tmp):
182	<	raise CrabException("Additional input file not found: "+tmp)
183	<	tmp=string.strip(tmp)
184	<	self.additional_inbox_files.append(tmp)
181	>	tmp = string.strip(tmp)
182	>	dirname = ''
183	>	if not tmp[0]=="/": dirname = "."
184	>	files = []
185	>	if string.find(tmp,"*")>-1:
186	>	files = glob.glob(os.path.join(dirname, tmp))
187	>	if len(files)==0:
188	>	raise CrabException("No additional input file found with this pattern: "+tmp)
189	>	else:
190	>	files.append(tmp)
191	>	for file in files:
192	>	if not os.path.exists(file):
193	>	raise CrabException("Additional input file not found: "+file)
194	>	pass
195	>	# fname = string.split(file, '/')[-1]
196	>	# storedFile = common.work_space.pathForTgz()+'share/'+fname
197	>	# shutil.copyfile(file, storedFile)
198	>	self.additional_inbox_files.append(string.strip(file))
199		pass
200		pass
201	+	common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
202		except KeyError:
203		pass
204
205	+	# files per job
206		try:
207	<	self.filesPerJob = int(cfg_params['CMSSW.files_per_jobs']) #Daniele
207	>	if (cfg_params['CMSSW.files_per_jobs']):
208	>	raise CrabException("files_per_jobs no longer supported.  Quitting.")
209		except KeyError:
210	<	self.filesPerJob = 1
210	>	pass
211
212	<	## Max event   will be total_number_of_events ???  Daniele
212	>	## Events per job
213		try:
214	<	self.maxEv = cfg_params['CMSSW.event_per_job']
214	>	self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
215	>	self.selectEventsPerJob = 1
216		except KeyError:
217	<	self.maxEv = "-1"
218	<	##
217	>	self.eventsPerJob = -1
218	>	self.selectEventsPerJob = 0
219	>
220	>	## number of jobs
221	>	try:
222	>	self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
223	>	self.selectNumberOfJobs = 1
224	>	except KeyError:
225	>	self.theNumberOfJobs = 0
226	>	self.selectNumberOfJobs = 0
227	>
228		try:
229		self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
230	+	self.selectTotalNumberEvents = 1
231		except KeyError:
232	<	msg = 'Must define total_number_of_events'
233	<	raise CrabException(msg)
234	<
235	<	CEBlackList = []
232	>	self.total_number_of_events = 0
233	>	self.selectTotalNumberEvents = 0
234	>
235	>	if self.pset != None: #CarlosDaniele
236	>	if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
237	>	msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
238	>	raise CrabException(msg)
239	>	else:
240	>	if (self.selectNumberOfJobs == 0):
241	>	msg = 'Must specify  number_of_jobs.'
242	>	raise CrabException(msg)
243	>
244	>	## source seed for pythia
245		try:
246	<	tmpBad = string.split(cfg_params['EDG.ce_black_list'],',')
164	<	for tmp in tmpBad:
165	<	tmp=string.strip(tmp)
166	<	CEBlackList.append(tmp)
246	>	self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
247		except KeyError:
248	<	pass
248	>	self.sourceSeed = None
249	>	common.logger.debug(5,"No seed given")
250
251	<	self.reCEBlackList=[]
252	<	for bad in CEBlackList:
172	<	self.reCEBlackList.append(re.compile( bad ))
173	<
174	<	common.logger.debug(5,'CEBlackList: '+str(CEBlackList))
175	<
176	<	CEWhiteList = []
177	<	try:
178	<	tmpGood = string.split(cfg_params['EDG.ce_white_list'],',')
179	<	for tmp in tmpGood:
180	<	tmp=string.strip(tmp)
181	<	CEWhiteList.append(tmp)
251	>	try:
252	>	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
253		except KeyError:
254	<	pass
254	>	self.sourceSeedVtx = None
255	>	common.logger.debug(5,"No vertex seed given")
256
257	<	#print 'CEWhiteList: ',CEWhiteList
258	<	self.reCEWhiteList=[]
259	<	for Good in CEWhiteList:
260	<	self.reCEWhiteList.append(re.compile( Good ))
257	>	try:
258	>	self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
259	>	except KeyError:
260	>	self.sourceSeedG4 = None
261	>	common.logger.debug(5,"No g4 sim hits seed given")
262
263	<	common.logger.debug(5,'CEWhiteList: '+str(CEWhiteList))
263	>	try:
264	>	self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
265	>	except KeyError:
266	>	self.sourceSeedMix = None
267	>	common.logger.debug(5,"No mix seed given")
268
269	<	self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
269	>	try:
270	>	self.firstRun = int(cfg_params['CMSSW.first_run'])
271	>	except KeyError:
272	>	self.firstRun = None
273	>	common.logger.debug(5,"No first run given")
274	>	if self.pset != None: #CarlosDaniele
275	>	ver = string.split(self.version,"_")
276	>	if (int(ver[1])>=1 and int(ver[2])>=5):
277	>	import PsetManipulator150 as pp
278	>	else:
279	>	import PsetManipulator as pp
280	>	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
281
282		#DBSDLS-start
283		## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
284		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
285		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
286	+	self.jobDestination=[]  # Site destination(s) for each job (list of lists)
287		## Perform the data location and discovery (based on DBS/DLS)
288	<	self.DataDiscoveryAndLocation(cfg_params)
288	>	## SL: Don't if NONE is specified as input (pythia use case)
289	>	blockSites = {}
290	>	if self.datasetPath:
291	>	blockSites = self.DataDiscoveryAndLocation(cfg_params)
292		#DBSDLS-end
293
294		self.tgzNameWithPath = self.getTarBall(self.executable)
295	+
296	+	## Select Splitting
297	+	if self.selectNoInput:
298	+	if self.pset == None: #CarlosDaniele
299	+	self.jobSplittingForScript()
300	+	else:
301	+	self.jobSplittingNoInput()
302	+	else:
303	+	self.jobSplittingByBlocks(blockSites)
304
305	<	self.jobSplitting()  #Daniele job Splitting
306	<	self.PsetEdit.maxEvent(self.maxEv) #Daniele
307	<	self.PsetEdit.inputModule("INPUT") #Daniele
308	<	self.PsetEdit.psetWriter(self.configFilename())
309	<
310	<
305	>	# modify Pset
306	>	if self.pset != None: #CarlosDaniele
307	>	try:
308	>	if (self.datasetPath): # standard job
309	>	# allow to processa a fraction of events in a file
310	>	PsetEdit.inputModule("INPUT")
311	>	PsetEdit.maxEvent("INPUTMAXEVENTS")
312	>	PsetEdit.skipEvent("INPUTSKIPEVENTS")
313	>	else:  # pythia like job
314	>	PsetEdit.maxEvent(self.eventsPerJob)
315	>	if (self.firstRun):
316	>	PsetEdit.pythiaFirstRun("INPUTFIRSTRUN")  #First Run
317	>	if (self.sourceSeed) :
318	>	PsetEdit.pythiaSeed("INPUT")
319	>	if (self.sourceSeedVtx) :
320	>	PsetEdit.vtxSeed("INPUTVTX")
321	>	if (self.sourceSeedG4) :
322	>	PsetEdit.g4Seed("INPUTG4")
323	>	if (self.sourceSeedMix) :
324	>	PsetEdit.mixSeed("INPUTMIX")
325	>	# add FrameworkJobReport to parameter-set
326	>	PsetEdit.addCrabFJR(self.fjrFileName)
327	>	PsetEdit.psetWriter(self.configFilename())
328	>	except:
329	>	msg='Error while manipuliating ParameterSet: exiting...'
330	>	raise CrabException(msg)
331
332		def DataDiscoveryAndLocation(self, cfg_params):
333
334	+	import DataDiscovery
335	+	import DataDiscovery_DBS2
336	+	import DataLocation
337		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
338
215	–	#datasetPath = "/"+self.owner+"/"+self.dataTiers[0]+"/"+self.dataset
216	–
339		datasetPath=self.datasetPath
340
219	–	## TODO
220	–	dataTiersList = ""
221	–	dataTiers = dataTiersList.split(',')
222	–
341		## Contact the DBS
342	+	common.logger.message("Contacting Data Discovery Services ...")
343		try:
344	<	self.pubdata=DataDiscovery_EDM.DataDiscovery_EDM(datasetPath, dataTiers, cfg_params)
344	>
345	>	if self.use_dbs_1 == 1 :
346	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
347	>	else :
348	>	self.pubdata=DataDiscovery_DBS2.DataDiscovery_DBS2(datasetPath, cfg_params)
349		self.pubdata.fetchDBSInfo()
350
351	<	except DataDiscovery_EDM.NotExistingDatasetError, ex :
351	>	except DataDiscovery.NotExistingDatasetError, ex :
352		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
353		raise CrabException(msg)
354	<
355	<	except DataDiscovery_EDM.NoDataTierinProvenanceError, ex :
354	>	except DataDiscovery.NoDataTierinProvenanceError, ex :
355	>	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
356	>	raise CrabException(msg)
357	>	except DataDiscovery.DataDiscoveryError, ex:
358	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
359	>	raise CrabException(msg)
360	>	except DataDiscovery_DBS2.NotExistingDatasetError_DBS2, ex :
361	>	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
362	>	raise CrabException(msg)
363	>	except DataDiscovery_DBS2.NoDataTierinProvenanceError_DBS2, ex :
364		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
365		raise CrabException(msg)
366	<	except DataDiscovery_EDM.DataDiscoveryError, ex:
367	<	msg = 'ERROR ***: failed Data Discovery in DBS  %s'%ex.getErrorMessage()
366	>	except DataDiscovery_DBS2.DataDiscoveryError_DBS2, ex:
367	>	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
368		raise CrabException(msg)
369
370	<	## get list of all required data in the form of dbs paths  (dbs path = /dataset/datatier/owner)
371	<	## self.DBSPaths=self.pubdata.getDBSPaths()
372	<	common.logger.message("Required data are :"+self.datasetPath)
242	<
243	<	filesbyblock=self.pubdata.getFiles()
244	<	self.AllInputFiles=filesbyblock.values()
245	<	self.files = self.AllInputFiles
246	<
247	<	## TEMP
248	<	#    self.filesTmp = filesbyblock.values()
249	<	#    self.files = []
250	<	#    locPath='rfio:cmsbose2.bo.infn.it:/flatfiles/SE00/cms/fanfani/ProdTest/'
251	<	#    locPath=''
252	<	#    tmp = []
253	<	#    for file in self.filesTmp[0]:
254	<	#        tmp.append(locPath+file)
255	<	#    self.files.append(tmp)
256	<	## END TEMP
370	>	self.filesbyblock=self.pubdata.getFiles()
371	>	self.eventsbyblock=self.pubdata.getEventsPerBlock()
372	>	self.eventsbyfile=self.pubdata.getEventsPerFile()
373
374		## get max number of events
259	–	#common.logger.debug(10,"number of events for primary fileblocks %i"%self.pubdata.getMaxEvents())
375		self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
261	–	common.logger.message("\nThe number of available events is %s"%self.maxEvents)
376
377		## Contact the DLS and build a list of sites hosting the fileblocks
378		try:
379	<	dataloc=DataLocation_EDM.DataLocation_EDM(filesbyblock.keys(),cfg_params)
379	>	dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
380		dataloc.fetchDLSInfo()
381	<	except DataLocation_EDM.DataLocationError , ex:
381	>	except DataLocation.DataLocationError , ex:
382		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
383		raise CrabException(msg)
384
271	–	allsites=dataloc.getSites()
272	–	common.logger.debug(5,"sites are %s"%allsites)
273	–	sites=self.checkBlackList(allsites)
274	–	common.logger.debug(5,"sites are (after black list) %s"%sites)
275	–	sites=self.checkWhiteList(sites)
276	–	common.logger.debug(5,"sites are (after white list) %s"%sites)
277	–
278	–	if len(sites)==0:
279	–	msg = 'No sites hosting all the needed data! Exiting... '
280	–	raise CrabException(msg)
385
386	<	common.logger.message("List of Sites hosting the data : "+str(sites))
387	<	common.logger.debug(6, "List of Sites: "+str(sites))
388	<	common.analisys_common_info['sites']=sites    ## used in SchedulerEdg.py in createSchScript
389	<	self.setParam_('TargetCE', ','.join(sites))
390	<	return
386	>	sites = dataloc.getSites()
387	>	allSites = []
388	>	listSites = sites.values()
389	>	for listSite in listSites:
390	>	for oneSite in listSite:
391	>	allSites.append(oneSite)
392	>	allSites = self.uniquelist(allSites)
393	>
394	>	# screen output
395	>	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
396	>
397	>	return sites
398
399	<	def jobSplitting(self):
399	>	def jobSplittingByBlocks(self, blockSites):
400		"""
401	<	first implemntation for job splitting
402	<	"""
403	<	#  print 'eventi totali '+str(self.maxEvents)
404	<	#  print 'eventi totali richiesti dallo user '+str(self.total_number_of_events)
405	<	#print 'files per job '+str(self.filesPerJob)
406	<	common.logger.message('Required '+str(self.filesPerJob)+' files per job ')
407	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
408	<
409	<	## TODO: SL need to have (from DBS) a detailed list of how many events per each file
410	<	n_tot_files = (len(self.files[0]))
411	<	## SL: this is wrong if the files have different number of events
412	<	evPerFile = int(self.maxEvents)/n_tot_files
413	<
414	<	common.logger.debug(5,'Events per File '+str(evPerFile))
415	<
416	<	## if asked to process all events, do it
417	<	if self.total_number_of_events == -1:
418	<	self.total_number_of_events=self.maxEvents
419	<	self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob)
420	<	common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for all available events '+str(self.total_number_of_events)+' events')
421	<
401	>	Perform job splitting. Jobs run over an integer number of files
402	>	and no more than one block.
403	>	ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
404	>	REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
405	>	self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
406	>	self.maxEvents, self.filesbyblock
407	>	SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
408	>	self.total_number_of_jobs - Total # of jobs
409	>	self.list_of_args - File(s) job will run on (a list of lists)
410	>	"""
411	>
412	>	# ---- Handle the possible job splitting configurations ---- #
413	>	if (self.selectTotalNumberEvents):
414	>	totalEventsRequested = self.total_number_of_events
415	>	if (self.selectEventsPerJob):
416	>	eventsPerJobRequested = self.eventsPerJob
417	>	if (self.selectNumberOfJobs):
418	>	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
419	>
420	>	# If user requested all the events in the dataset
421	>	if (totalEventsRequested == -1):
422	>	eventsRemaining=self.maxEvents
423	>	# If user requested more events than are in the dataset
424	>	elif (totalEventsRequested > self.maxEvents):
425	>	eventsRemaining = self.maxEvents
426	>	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
427	>	# If user requested less events than are in the dataset
428		else:
429	<	self.total_number_of_files = int(self.total_number_of_events/evPerFile)
430	<	## SL: if ask for less event than what is computed to be available on a
431	<	##     file, process the first file anyhow.
432	<	if self.total_number_of_files == 0:
433	<	self.total_number_of_files = self.total_number_of_files + 1
429	>	eventsRemaining = totalEventsRequested
430	>
431	>	# If user requested more events per job than are in the dataset
432	>	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
433	>	eventsPerJobRequested = self.maxEvents
434	>
435	>	# For user info at end
436	>	totalEventCount = 0
437
438	<	common.logger.debug(5,'N files  '+str(self.total_number_of_files))
438	>	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
439	>	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
440	>
441	>	if (self.selectNumberOfJobs):
442	>	common.logger.message("May not create the exact number_of_jobs requested.")
443	>
444	>	if ( self.ncjobs == 'all' ) :
445	>	totalNumberOfJobs = 999999999
446	>	else :
447	>	totalNumberOfJobs = self.ncjobs
448	>
449
450	<	check = 0
450	>	blocks = blockSites.keys()
451	>	blockCount = 0
452	>	# Backup variable in case self.maxEvents counted events in a non-included block
453	>	numBlocksInDataset = len(blocks)
454	>
455	>	jobCount = 0
456	>	list_of_lists = []
457	>
458	>	# list tracking which jobs are in which jobs belong to which block
459	>	jobsOfBlock = {}
460	>
461	>	# ---- Iterate over the blocks in the dataset until ---- #
462	>	# ---- we've met the requested total # of events    ---- #
463	>	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
464	>	block = blocks[blockCount]
465	>	blockCount += 1
466	>	if block not in jobsOfBlock.keys() :
467	>	jobsOfBlock[block] = []
468	>
469	>	if self.eventsbyblock.has_key(block) :
470	>	numEventsInBlock = self.eventsbyblock[block]
471	>	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
472	>
473	>	files = self.filesbyblock[block]
474	>	numFilesInBlock = len(files)
475	>	if (numFilesInBlock <= 0):
476	>	continue
477	>	fileCount = 0
478	>
479	>	# ---- New block => New job ---- #
480	>	parString = "\\{"
481	>	# counter for number of events in files currently worked on
482	>	filesEventCount = 0
483	>	# flag if next while loop should touch new file
484	>	newFile = 1
485	>	# job event counter
486	>	jobSkipEventCount = 0
487
488	<	## Compute the number of jobs
489	<	#self.total_number_of_jobs = int(n_tot_files)*1/int(self.filesPerJob)
490	<	self.total_number_of_jobs = int(self.total_number_of_files/self.filesPerJob)
491	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
488	>	# ---- Iterate over the files in the block until we've met the requested ---- #
489	>	# ---- total # of events or we've gone over all the files in this block  ---- #
490	>	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
491	>	file = files[fileCount]
492	>	if newFile :
493	>	try:
494	>	numEventsInFile = self.eventsbyfile[file]
495	>	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
496	>	# increase filesEventCount
497	>	filesEventCount += numEventsInFile
498	>	# Add file to current job
499	>	parString += '\\\"' + file + '\\\"\,'
500	>	newFile = 0
501	>	except KeyError:
502	>	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
503	>
504	>
505	>	# if less events in file remain than eventsPerJobRequested
506	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
507	>	# if last file in block
508	>	if ( fileCount == numFilesInBlock-1 ) :
509	>	# end job using last file, use remaining events in block
510	>	# close job and touch new file
511	>	fullString = parString[:-2]
512	>	fullString += '\\}'
513	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
514	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
515	>	self.jobDestination.append(blockSites[block])
516	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
517	>	# fill jobs of block dictionary
518	>	jobsOfBlock[block].append(jobCount+1)
519	>	# reset counter
520	>	jobCount = jobCount + 1
521	>	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
522	>	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
523	>	jobSkipEventCount = 0
524	>	# reset file
525	>	parString = "\\{"
526	>	filesEventCount = 0
527	>	newFile = 1
528	>	fileCount += 1
529	>	else :
530	>	# go to next file
531	>	newFile = 1
532	>	fileCount += 1
533	>	# if events in file equal to eventsPerJobRequested
534	>	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
535	>	# close job and touch new file
536	>	fullString = parString[:-2]
537	>	fullString += '\\}'
538	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
539	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
540	>	self.jobDestination.append(blockSites[block])
541	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
542	>	jobsOfBlock[block].append(jobCount+1)
543	>	# reset counter
544	>	jobCount = jobCount + 1
545	>	totalEventCount = totalEventCount + eventsPerJobRequested
546	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
547	>	jobSkipEventCount = 0
548	>	# reset file
549	>	parString = "\\{"
550	>	filesEventCount = 0
551	>	newFile = 1
552	>	fileCount += 1
553	>
554	>	# if more events in file remain than eventsPerJobRequested
555	>	else :
556	>	# close job but don't touch new file
557	>	fullString = parString[:-2]
558	>	fullString += '\\}'
559	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
560	>	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
561	>	self.jobDestination.append(blockSites[block])
562	>	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
563	>	jobsOfBlock[block].append(jobCount+1)
564	>	# increase counter
565	>	jobCount = jobCount + 1
566	>	totalEventCount = totalEventCount + eventsPerJobRequested
567	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
568	>	# calculate skip events for last file
569	>	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
570	>	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
571	>	# remove all but the last file
572	>	filesEventCount = self.eventsbyfile[file]
573	>	parString = "\\{"
574	>	parString += '\\\"' + file + '\\\"\,'
575	>	pass # END if
576	>	pass # END while (iterate over files in the block)
577	>	pass # END while (iterate over blocks in the dataset)
578	>	self.ncjobs = self.total_number_of_jobs = jobCount
579	>	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
580	>	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
581	>	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
582	>
583	>	# screen output
584	>	screenOutput = "List of jobs and available destination sites:\n\n"
585	>
586	>	blockCounter = 0
587	>	for block in blocks:
588	>	if block in jobsOfBlock.keys() :
589	>	blockCounter += 1
590	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
591	>
592	>	common.logger.message(screenOutput)
593	>
594	>	self.list_of_args = list_of_lists
595	>	return
596	>
597	>	def jobSplittingNoInput(self):
598	>	"""
599	>	Perform job splitting based on number of event per job
600	>	"""
601	>	common.logger.debug(5,'Splitting per events')
602	>	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
603	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
604	>	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
605
606	<	## is there any remainder?
607	<	check = int(self.total_number_of_files) - (int(self.total_number_of_jobs)*self.filesPerJob)
606	>	if (self.total_number_of_events < 0):
607	>	msg='Cannot split jobs per Events with "-1" as total number of events'
608	>	raise CrabException(msg)
609
610	<	common.logger.debug(5,'Check  '+str(check))
610	>	if (self.selectEventsPerJob):
611	>	if (self.selectTotalNumberEvents):
612	>	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
613	>	elif(self.selectNumberOfJobs) :
614	>	self.total_number_of_jobs =self.theNumberOfJobs
615	>	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
616	>
617	>	elif (self.selectNumberOfJobs) :
618	>	self.total_number_of_jobs = self.theNumberOfJobs
619	>	self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
620	>
621	>	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
622
623	<	if check > 0:
624	<	self.total_number_of_jobs =  self.total_number_of_jobs + 1
334	<	common.logger.message('Warning: last job will be created with '+str(check)+' files')
623	>	# is there any remainder?
624	>	check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob)
625
626	<	common.logger.message(str(self.total_number_of_jobs)+' jobs will be created for a total of '+str((self.total_number_of_jobs-1)*self.filesPerJob*evPerFile + check*evPerFile)+' events')
626	>	common.logger.debug(5,'Check  '+str(check))
627	>
628	>	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created, each for '+str(self.eventsPerJob)+' for a total of '+str(self.total_number_of_jobs*self.eventsPerJob)+' events')
629	>	if check > 0:
630	>	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
631	>
632	>	# argument is seed number.$i
633	>	self.list_of_args = []
634	>	for i in range(self.total_number_of_jobs):
635	>	## Since there is no input, any site is good
636	>	# self.jobDestination.append(["Any"])
637	>	self.jobDestination.append([""]) #must be empty to write correctly the xml
638	>	args=[]
639	>	if (self.firstRun):
640	>	## pythia first run
641	>	#self.list_of_args.append([(str(self.firstRun)+str(i))])
642	>	args.append(str(self.firstRun)+str(i))
643	>	else:
644	>	## no first run
645	>	#self.list_of_args.append([str(i)])
646	>	args.append(str(i))
647	>	if (self.sourceSeed):
648	>	args.append(str(self.sourceSeed)+str(i))
649	>	if (self.sourceSeedVtx):
650	>	## + vtx random seed
651	>	args.append(str(self.sourceSeedVtx)+str(i))
652	>	if (self.sourceSeedG4):
653	>	## + G4 random seed
654	>	args.append(str(self.sourceSeedG4)+str(i))
655	>	if (self.sourceSeedMix):
656	>	## + Mix random seed
657	>	args.append(str(self.sourceSeedMix)+str(i))
658	>	pass
659		pass
660	+	self.list_of_args.append(args)
661	+	pass
662	+
663	+	# print self.list_of_args
664
665	<	list_of_lists = []
340	<	for i in xrange(0, int(n_tot_files), self.filesPerJob):
341	<	list_of_lists.append(self.files[0][i: i+self.filesPerJob])
665	>	return
666
667	<	self.list_of_files = list_of_lists
668	<
667	>
668	>	def jobSplittingForScript(self):#CarlosDaniele
669	>	"""
670	>	Perform job splitting based on number of job
671	>	"""
672	>	common.logger.debug(5,'Splitting per job')
673	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
674	>
675	>	self.total_number_of_jobs = self.theNumberOfJobs
676	>
677	>	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
678	>
679	>	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
680	>
681	>	# argument is seed number.$i
682	>	self.list_of_args = []
683	>	for i in range(self.total_number_of_jobs):
684	>	## Since there is no input, any site is good
685	>	# self.jobDestination.append(["Any"])
686	>	self.jobDestination.append([""])
687	>	## no random seed
688	>	self.list_of_args.append([str(i)])
689		return
690
691		def split(self, jobParams):
#	Line 349 | Line 693 | class Cmssw(JobType):
693		common.jobDB.load()
694		#### Fabio
695		njobs = self.total_number_of_jobs
696	<	filelist = self.list_of_files
696	>	arglist = self.list_of_args
697		# create the empty structure
698		for i in range(njobs):
699		jobParams.append("")
700
701		for job in range(njobs):
702	<	jobParams[job] = filelist[job]
702	>	jobParams[job] = arglist[job]
703	>	# print str(arglist[job])
704	>	# print jobParams[job]
705		common.jobDB.setArguments(job, jobParams[job])
706	+	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
707	+	common.jobDB.setDestination(job, self.jobDestination[job])
708
709		common.jobDB.save()
710		return
711
712		def getJobTypeArguments(self, nj, sched):
713	<	params = common.jobDB.arguments(nj)
714	<	#print params
715	<	parString = "\\{"
716	<
369	<	for i in range(len(params) - 1):
370	<	parString += '\\\"' + params[i] + '\\\"\,'
371	<
372	<	parString += '\\\"' + params[len(params) - 1] + '\\\"\\}'
373	<	return parString
713	>	result = ''
714	>	for i in common.jobDB.arguments(nj):
715	>	result=result+str(i)+" "
716	>	return result
717
718		def numberOfJobs(self):
719		# Fabio
377	–
720		return self.total_number_of_jobs
379	–
380	–
381	–
382	–	def checkBlackList(self, allSites):
383	–	if len(self.reCEBlackList)==0: return allSites
384	–	sites = []
385	–	for site in allSites:
386	–	common.logger.debug(10,'Site '+site)
387	–	good=1
388	–	for re in self.reCEBlackList:
389	–	if re.search(site):
390	–	common.logger.message('CE in black list, skipping site '+site)
391	–	good=0
392	–	pass
393	–	if good: sites.append(site)
394	–	if len(sites) == 0:
395	–	common.logger.debug(3,"No sites found after BlackList")
396	–	return sites
397	–
398	–	def checkWhiteList(self, allSites):
399	–
400	–	if len(self.reCEWhiteList)==0: return allSites
401	–	sites = []
402	–	for site in allSites:
403	–	good=0
404	–	for re in self.reCEWhiteList:
405	–	if re.search(site):
406	–	common.logger.debug(5,'CE in white list, adding site '+site)
407	–	good=1
408	–	if not good: continue
409	–	sites.append(site)
410	–	if len(sites) == 0:
411	–	common.logger.message("No sites found after WhiteList\n")
412	–	else:
413	–	common.logger.debug(5,"Selected sites via WhiteList are "+str(sites)+"\n")
414	–	return sites
721
722		def getTarBall(self, exe):
723		"""
#	Line 419 | Line 725 | class Cmssw(JobType):
725		"""
726
727		# if it exist, just return it
728	<	self.tgzNameWithPath = common.work_space.shareDir()+self.tgz_name
728	>	#
729	>	# Marco. Let's start to use relative path for Boss XML files
730	>	#
731	>	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
732		if os.path.exists(self.tgzNameWithPath):
733		return self.tgzNameWithPath
734
#	Line 433 | Line 742 | class Cmssw(JobType):
742		# First of all declare the user Scram area
743		swArea = self.scram.getSWArea_()
744		#print "swArea = ", swArea
745	<	swVersion = self.scram.getSWVersion()
746	<	#print "swVersion = ", swVersion
745	>	# swVersion = self.scram.getSWVersion()
746	>	# print "swVersion = ", swVersion
747		swReleaseTop = self.scram.getReleaseTop_()
748		#print "swReleaseTop = ", swReleaseTop
749
#	Line 442 | Line 751 | class Cmssw(JobType):
751		if swReleaseTop == '' or swArea == swReleaseTop:
752		return
753
754	<	filesToBeTarred = []
755	<	## First find the executable
756	<	if (self.executable != ''):
757	<	exeWithPath = self.scram.findFile_(executable)
758	<	#           print exeWithPath
759	<	if ( not exeWithPath ):
760	<	raise CrabException('User executable '+executable+' not found')
761	<
762	<	## then check if it's private or not
763	<	if exeWithPath.find(swReleaseTop) == -1:
764	<	# the exe is private, so we must ship
765	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
766	<	path = swArea+'/'
767	<	exe = string.replace(exeWithPath, path,'')
768	<	filesToBeTarred.append(exe)
769	<	pass
770	<	else:
771	<	# the exe is from release, we'll find it on WN
772	<	pass
773	<
774	<	## Now get the libraries: only those in local working area
775	<	libDir = 'lib'
776	<	lib = swArea+'/' +libDir
777	<	common.logger.debug(5,"lib "+lib+" to be tarred")
778	<	if os.path.exists(lib):
779	<	filesToBeTarred.append(libDir)
780	<
781	<	## Now check if module dir is present
782	<	moduleDir = 'module'
783	<	if os.path.isdir(swArea+'/'+moduleDir):
784	<	filesToBeTarred.append(moduleDir)
785	<
786	<	## Now check if the Data dir is present
787	<	dataDir = 'src/Data/'
788	<	if os.path.isdir(swArea+'/'+dataDir):
789	<	filesToBeTarred.append(dataDir)
790	<
791	<	## Create the tar-ball
792	<	if len(filesToBeTarred)>0:
793	<	cwd = os.getcwd()
794	<	os.chdir(swArea)
795	<	tarcmd = 'tar zcvf ' + self.tgzNameWithPath + ' '
796	<	for line in filesToBeTarred:
797	<	tarcmd = tarcmd + line + ' '
798	<	cout = runCommand(tarcmd)
799	<	if not cout:
800	<	raise CrabException('Could not create tar-ball')
801	<	os.chdir(cwd)
802	<	else:
803	<	common.logger.debug(5,"No files to be to be tarred")
754	>	import tarfile
755	>	try: # create tar ball
756	>	tar = tarfile.open(self.tgzNameWithPath, "w:gz")
757	>	## First find the executable
758	>	if (self.executable != ''):
759	>	exeWithPath = self.scram.findFile_(executable)
760	>	if ( not exeWithPath ):
761	>	raise CrabException('User executable '+executable+' not found')
762	>
763	>	## then check if it's private or not
764	>	if exeWithPath.find(swReleaseTop) == -1:
765	>	# the exe is private, so we must ship
766	>	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
767	>	path = swArea+'/'
768	>	# distinguish case when script is in user project area or given by full path somewhere else
769	>	if exeWithPath.find(path) >= 0 :
770	>	exe = string.replace(exeWithPath, path,'')
771	>	tar.add(path+exe,os.path.basename(executable))
772	>	else :
773	>	tar.add(exeWithPath,os.path.basename(executable))
774	>	pass
775	>	else:
776	>	# the exe is from release, we'll find it on WN
777	>	pass
778	>
779	>	## Now get the libraries: only those in local working area
780	>	libDir = 'lib'
781	>	lib = swArea+'/' +libDir
782	>	common.logger.debug(5,"lib "+lib+" to be tarred")
783	>	if os.path.exists(lib):
784	>	tar.add(lib,libDir)
785	>
786	>	## Now check if module dir is present
787	>	moduleDir = 'module'
788	>	module = swArea + '/' + moduleDir
789	>	if os.path.isdir(module):
790	>	tar.add(module,moduleDir)
791	>
792	>	## Now check if any data dir(s) is present
793	>	swAreaLen=len(swArea)
794	>	for root, dirs, files in os.walk(swArea):
795	>	if "data" in dirs:
796	>	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
797	>	tar.add(root+"/data",root[swAreaLen:]+"/data")
798	>
799	>	## Add ProdAgent dir to tar
800	>	paDir = 'ProdAgentApi'
801	>	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
802	>	if os.path.isdir(pa):
803	>	tar.add(pa,paDir)
804	>
805	>	### FEDE FOR DBS PUBLICATION
806	>	## Add PRODCOMMON dir to tar
807	>	prodcommonDir = 'ProdCommon'
808	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
809	>	if os.path.isdir(prodcommonPath):
810	>	tar.add(prodcommonPath,prodcommonDir)
811	>	#############################
812	>
813	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
814	>	tar.close()
815	>	except :
816	>	raise CrabException('Could not create tar-ball')
817	>
818	>	## check for tarball size
819	>	tarballinfo = os.stat(self.tgzNameWithPath)
820	>	if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
821	>	raise CrabException('Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + str(self.MaxTarBallSize) + ' MB input sandbox limit and not supported by the used GRID submission system. Please make sure that no unnecessary files are in all data directories in your local CMSSW project area as they are automatically packed into the input sandbox.')
822	>
823	>	## create tar-ball with ML stuff
824	>	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
825	>	try:
826	>	tar = tarfile.open(self.MLtgzfile, "w:gz")
827	>	path=os.environ['CRABDIR'] + '/python/'
828	>	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
829	>	tar.add(path+file,file)
830	>	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
831	>	tar.close()
832	>	except :
833	>	raise CrabException('Could not create ML files tar-ball')
834
835		return
836
837	+	def additionalInputFileTgz(self):
838	+	"""
839	+	Put all additional files into a tar ball and return its name
840	+	"""
841	+	import tarfile
842	+	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
843	+	tar = tarfile.open(tarName, "w:gz")
844	+	for file in self.additional_inbox_files:
845	+	tar.add(file,string.split(file,'/')[-1])
846	+	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
847	+	tar.close()
848	+	return tarName
849	+
850		def wsSetupEnvironment(self, nj):
851		"""
852		Returns part of a job script which prepares
#	Line 506 | Line 858 | class Cmssw(JobType):
858		## OLI_Daniele at this level  middleware already known
859
860		txt += 'if [ $middleware == LCG ]; then \n'
861	+	txt += '    echo "### First set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
862	+	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
863	+	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
864		txt += self.wsSetupCMSLCGEnvironment_()
865		txt += 'elif [ $middleware == OSG ]; then\n'
866	<	txt += '    time=`date -u +"%s"`\n'
867	<	txt += '    WORKING_DIR=$OSG_WN_TMP/cms_$time\n'
513	<	txt += '    echo "Creating working directory: $WORKING_DIR"\n'
514	<	txt += '    /bin/mkdir -p $WORKING_DIR\n'
866	>	txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
867	>	txt += '    echo "Created working directory: $WORKING_DIR"\n'
868		txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
869		txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
870	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
871	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
872	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
870	>	txt += '    echo "JOB_EXIT_STATUS = 10016"\n'
871	>	txt += '    echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
872	>	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
873	>	txt += '        rm -f $RUNTIME_AREA/$repo \n'
874	>	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
875	>	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
876		txt += '        exit 1\n'
877		txt += '    fi\n'
878		txt += '\n'
879		txt += '    echo "Change to working directory: $WORKING_DIR"\n'
880		txt += '    cd $WORKING_DIR\n'
881		txt += self.wsSetupCMSOSGEnvironment_()
882	+	txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
883	+	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
884		txt += 'fi\n'
885
886		# Prepare JobType-specific part
#	Line 536 | Line 894 | class Cmssw(JobType):
894		txt += '   echo "JOB_EXIT_STATUS = 10034"\n'
895		txt += '   echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
896		txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
897	+	txt += '   rm -f $RUNTIME_AREA/$repo \n'
898	+	txt += '   echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
899	+	txt += '   echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
900		## OLI_Daniele
901		txt += '    if [ $middleware == OSG ]; then \n'
902		txt += '        echo "Remove working directory: $WORKING_DIR"\n'
903		txt += '        cd $RUNTIME_AREA\n'
904		txt += '        /bin/rm -rf $WORKING_DIR\n'
905		txt += '        if [ -d $WORKING_DIR ] ;then\n'
906	<	txt += '            echo "SET_CMS_ENV 10018 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after CMSSW CMSSW_0_6_1 not found on `hostname`"\n'
907	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
908	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
909	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
906	>	txt += '            echo "SET_CMS_ENV 10018 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after CMSSW CMSSW_0_6_1 not found on `hostname`"\n'
907	>	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
908	>	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
909	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
910	>	txt += '            rm -f $RUNTIME_AREA/$repo \n'
911	>	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
912	>	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
913		txt += '        fi\n'
914		txt += '    fi \n'
915		txt += '   exit 1 \n'
916		txt += 'fi \n'
917		txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
918		txt += 'cd '+self.version+'\n'
919	+	########## FEDE FOR DBS2 ######################
920	+	txt += 'SOFTWARE_DIR=`pwd`\n'
921	+	txt += 'echo SOFTWARE_DIR=$SOFTWARE_DIR \n'
922	+	###############################################
923		### needed grep for bug in scramv1 ###
924	+	txt += scram+' runtime -sh\n'
925		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
926	+	txt += 'echo $PATH\n'
927
928		# Handle the arguments:
929		txt += "\n"
930		txt += "## number of arguments (first argument always jobnumber)\n"
931		txt += "\n"
932	<	txt += "narg=$#\n"
933	<	txt += "if [ $narg -lt 2 ]\n"
932	>	#        txt += "narg=$#\n"
933	>	txt += "if [ $nargs -lt 2 ]\n"
934		txt += "then\n"
935	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$narg+ \n"
935	>	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
936		txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
937		txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
938		txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
939	+	txt += '    rm -f $RUNTIME_AREA/$repo \n'
940	+	txt += '    echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
941	+	txt += '    echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
942		## OLI_Daniele
943		txt += '    if [ $middleware == OSG ]; then \n'
944		txt += '        echo "Remove working directory: $WORKING_DIR"\n'
945		txt += '        cd $RUNTIME_AREA\n'
946		txt += '        /bin/rm -rf $WORKING_DIR\n'
947		txt += '        if [ -d $WORKING_DIR ] ;then\n'
948	<	txt += '            echo "SET_EXE_ENV 50114 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Too few arguments for CRAB job wrapper"\n'
949	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
950	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
951	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
948	>	txt += '            echo "SET_EXE_ENV 50114 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Too few arguments for CRAB job wrapper"\n'
949	>	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
950	>	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
951	>	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
952	>	txt += '            rm -f $RUNTIME_AREA/$repo \n'
953	>	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
954	>	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
955		txt += '        fi\n'
956		txt += '    fi \n'
957		txt += "    exit 1\n"
#	Line 584 | Line 960 | class Cmssw(JobType):
960
961		# Prepare job-specific part
962		job = common.job_list[nj]
963	<	pset = os.path.basename(job.configFilename())
964	<	txt += '\n'
965	<	txt += 'InputFiles=$2\n'
966	<	txt += 'echo "<$InputFiles>"\n'
967	<	#txt += 'echo sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' \n'
968	<	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset.cfg\n'
969	<	#txt += 'sed "s#{\'INPUT\'}#${InputFiles}#" $RUNTIME_AREA/'+pset+' > pset1.cfg\n'
963	>	### FEDE FOR DBS OUTPUT PUBLICATION
964	>	if (self.datasetPath):
965	>	txt += '\n'
966	>	txt += 'DatasetPath='+self.datasetPath+'\n'
967	>
968	>	datasetpath_split = self.datasetPath.split("/")
969	>
970	>	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
971	>	txt += 'DataTier='+datasetpath_split[2]+'\n'
972	>	#txt += 'ProcessedDataset='+datasetpath_split[3]+'\n'
973	>	txt += 'ApplicationFamily=cmsRun\n'
974	>
975	>	else:
976	>	txt += 'DatasetPath=MCDataTier\n'
977	>	txt += 'PrimaryDataset=null\n'
978	>	txt += 'DataTier=null\n'
979	>	#txt += 'ProcessedDataset=null\n'
980	>	txt += 'ApplicationFamily=MCDataTier\n'
981	>	if self.pset != None: #CarlosDaniele
982	>	pset = os.path.basename(job.configFilename())
983	>	txt += '\n'
984	>	txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
985	>	if (self.datasetPath): # standard job
986	>	#txt += 'InputFiles=$2\n'
987	>	txt += 'InputFiles=${args[1]}\n'
988	>	txt += 'MaxEvents=${args[2]}\n'
989	>	txt += 'SkipEvents=${args[3]}\n'
990	>	txt += 'echo "Inputfiles:<$InputFiles>"\n'
991	>	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
992	>	txt += 'echo "MaxEvents:<$MaxEvents>"\n'
993	>	txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
994	>	txt += 'echo "SkipEvents:<$SkipEvents>"\n'
995	>	txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
996	>	else:  # pythia like job
997	>	seedIndex=1
998	>	if (self.firstRun):
999	>	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
1000	>	txt += 'echo "FirstRun: <$FirstRun>"\n'
1001	>	txt += 'sed "s#\<INPUTFIRSTRUN\>#$FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1002	>	seedIndex=seedIndex+1
1003	>
1004	>	if (self.sourceSeed):
1005	>	txt += 'Seed=${args['+str(seedIndex)+']}\n'
1006	>	txt += 'sed "s#\<INPUT\>#$Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1007	>	seedIndex=seedIndex+1
1008	>	## the following seeds are not always present
1009	>	if (self.sourceSeedVtx):
1010	>	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
1011	>	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
1012	>	txt += 'sed "s#\<INPUTVTX\>#$VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1013	>	seedIndex += 1
1014	>	if (self.sourceSeedG4):
1015	>	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
1016	>	txt += 'echo "G4Seed: <$G4Seed>"\n'
1017	>	txt += 'sed "s#\<INPUTG4\>#$G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1018	>	seedIndex += 1
1019	>	if (self.sourceSeedMix):
1020	>	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
1021	>	txt += 'echo "MixSeed: <$mixSeed>"\n'
1022	>	txt += 'sed "s#\<INPUTMIX\>#$mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1023	>	seedIndex += 1
1024	>	pass
1025	>	pass
1026	>	txt += 'mv -f '+pset+' pset.cfg\n'
1027
1028		if len(self.additional_inbox_files) > 0:
1029	<	for file in self.additional_inbox_files:
1030	<	txt += 'if [ -e $RUNTIME_AREA/'+file+' ] ; then\n'
1031	<	txt += '   cp $RUNTIME_AREA/'+file+' .\n'
599	<	txt += '   chmod +x '+file+'\n'
600	<	txt += 'fi\n'
1029	>	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
1030	>	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
1031	>	txt += 'fi\n'
1032		pass
1033
1034	<	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
1035	<
1036	<	txt += '\n'
1037	<	txt += 'echo "***** cat pset.cfg *********"\n'
1038	<	txt += 'cat pset.cfg\n'
1039	<	txt += 'echo "****** end pset.cfg ********"\n'
1040	<	txt += '\n'
1041	<	# txt += 'echo "***** cat pset1.cfg *********"\n'
1042	<	# txt += 'cat pset1.cfg\n'
1043	<	# txt += 'echo "****** end pset1.cfg ********"\n'
1034	>	if self.pset != None: #CarlosDaniele
1035	>	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
1036	>
1037	>	txt += '\n'
1038	>	txt += 'echo "***** cat pset.cfg *********"\n'
1039	>	txt += 'cat pset.cfg\n'
1040	>	txt += 'echo "****** end pset.cfg ********"\n'
1041	>	txt += '\n'
1042	>	### FEDE FOR DBS OUTPUT PUBLICATION
1043	>	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
1044	>	txt += 'echo "PSETHASH = $PSETHASH" \n'
1045	>	##############
1046	>	txt += '\n'
1047	>	# txt += 'echo "***** cat pset1.cfg *********"\n'
1048	>	# txt += 'cat pset1.cfg\n'
1049	>	# txt += 'echo "****** end pset1.cfg ********"\n'
1050		return txt
1051
1052	<	def wsBuildExe(self, nj):
1052	>	def wsBuildExe(self, nj=0):
1053		"""
1054		Put in the script the commands to build an executable
1055		or a library.
#	Line 633 | Line 1070 | class Cmssw(JobType):
1070		txt += '       cd $RUNTIME_AREA\n'
1071		txt += '       /bin/rm -rf $WORKING_DIR\n'
1072		txt += '       if [ -d $WORKING_DIR ] ;then\n'
1073	<	txt += '        echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1074	<	txt += '        echo "JOB_EXIT_STATUS = 50999"\n'
1075	<	txt += '        echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1076	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1073	>	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1074	>	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
1075	>	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1076	>	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1077	>	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1078	>	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1079	>	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1080		txt += '       fi\n'
1081		txt += '   fi \n'
1082		txt += '   \n'
#	Line 644 | Line 1084 | class Cmssw(JobType):
1084		txt += 'else \n'
1085		txt += '   echo "Successful untar" \n'
1086		txt += 'fi \n'
1087	+	txt += '\n'
1088	+	txt += 'echo "Include ProdAgentApi and PRODCOMMON in PYTHONPATH"\n'
1089	+	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1090	+	#### FEDE FOR DBS OUTPUT PUBLICATION
1091	+	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
1092	+	#txt += '   export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon\n'
1093	+	#txt += '   export PYTHONPATH=ProdAgentApi\n'
1094	+	txt += 'else\n'
1095	+	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
1096	+	#txt += '   export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon:${PYTHONPATH}\n'
1097	+	#txt += '   export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
1098	+	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1099	+	###################
1100	+	txt += 'fi\n'
1101	+	txt += '\n'
1102	+
1103		pass
1104
1105		return txt
#	Line 655 | Line 1111 | class Cmssw(JobType):
1111		"""
1112
1113		def executableName(self):
1114	<	return self.executable
1114	>	if self.scriptExe: #CarlosDaniele
1115	>	return "sh "
1116	>	else:
1117	>	return self.executable
1118
1119		def executableArgs(self):
1120	<	return " -p pset.cfg"
1120	>	if self.scriptExe:#CarlosDaniele
1121	>	return   self.scriptExe + " $NJob"
1122	>	else:
1123	>	# if >= CMSSW_1_5_X, add -e
1124	>	version_array = self.scram.getSWVersion().split('_')
1125	>	major = 0
1126	>	minor = 0
1127	>	try:
1128	>	major = int(version_array[1])
1129	>	minor = int(version_array[2])
1130	>	except:
1131	>	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1132	>	raise CrabException(msg)
1133	>	if major >= 1 and minor >= 5 :
1134	>	return " -e -p pset.cfg"
1135	>	else:
1136	>	return " -p pset.cfg"
1137
1138		def inputSandbox(self, nj):
1139		"""
1140		Returns a list of filenames to be put in JDL input sandbox.
1141		"""
1142		inp_box = []
1143	<	# dict added to delete duplicate from input sandbox file list
1144	<	seen = {}
1143	>	# # dict added to delete duplicate from input sandbox file list
1144	>	# seen = {}
1145		## code
1146		if os.path.isfile(self.tgzNameWithPath):
1147		inp_box.append(self.tgzNameWithPath)
1148	+	if os.path.isfile(self.MLtgzfile):
1149	+	inp_box.append(self.MLtgzfile)
1150		## config
1151	<	inp_box.append(common.job_list[nj].configFilename())
1151	>	if not self.pset is None:
1152	>	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1153		## additional input files
1154	<	#for file in self.additional_inbox_files:
1155	<	#    inp_box.append(common.work_space.cwdDir()+file)
1154	>	tgz = self.additionalInputFileTgz()
1155	>	inp_box.append(tgz)
1156		return inp_box
1157
1158		def outputSandbox(self, nj):
#	Line 683 | Line 1161 | class Cmssw(JobType):
1161		"""
1162		out_box = []
1163
686	–	stdout=common.job_list[nj].stdout()
687	–	stderr=common.job_list[nj].stderr()
688	–
1164		## User Declared output files
1165	<	for out in self.output_file:
1165	>	for out in (self.output_file+self.output_file_sandbox):
1166		n_out = nj + 1
1167		out_box.append(self.numberFile_(out,str(n_out)))
1168		return out_box
694	–	return []
1169
1170		def prepareSteeringCards(self):
1171		"""
#	Line 707 | Line 1181 | class Cmssw(JobType):
1181		txt = '\n'
1182		txt += '# directory content\n'
1183		txt += 'ls \n'
1184	<	file_list = ''
1185	<	for fileWithSuffix in self.output_file:
1184	>
1185	>	for fileWithSuffix in (self.output_file+self.output_file_sandbox):
1186		output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
713	–	file_list=file_list+output_file_num+' '
1187		txt += '\n'
1188		txt += '# check output file\n'
1189	<	txt += 'ls '+fileWithSuffix+'\n'
1190	<	txt += 'exe_result=$?\n'
1191	<	txt += 'if [ $exe_result -ne 0 ] ; then\n'
1192	<	txt += '   echo "ERROR: No output file to manage"\n'
1193	<	txt += '   echo "JOB_EXIT_STATUS = $exe_result"\n'
1194	<	txt += '   echo "JobExitCode=60302" | tee -a $RUNTIME_AREA/$repo\n'
1195	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
1196	<	### OLI_DANIELE
1189	>	# txt += 'ls '+fileWithSuffix+'\n'
1190	>	# txt += 'ls_result=$?\n'
1191	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1192	>	###### FEDE FOR OUTPUT DATA PUBLICATION ########
1193	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1194	>	txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1195	>	################################################
1196	>	txt += 'else\n'
1197	>	txt += '    exit_status=60302\n'
1198	>	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1199	>	############# FEDE ADDED CHECK FOR OUTPUT #############
1200	>	## MATTY's FIX: the exit option was interrupting the execution
1201	>	if fileWithSuffix in self.output_file:
1202	>	txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1203	>	txt += '    output_exit_status=$exit_status\n'
1204	>	txt += '    # exit $exit_status\n'
1205	>	#######################################################
1206		if common.scheduler.boss_scheduler_name == 'condor_g':
1207		txt += '    if [ $middleware == OSG ]; then \n'
1208		txt += '        echo "prepare dummy output file"\n'
1209		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1210		txt += '    fi \n'
729	–	txt += 'else\n'
730	–	txt += '   cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1211		txt += 'fi\n'
1212	<
1212	>	file_list = []
1213	>	for fileWithSuffix in (self.output_file):
1214	>	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1215	>
1216	>	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1217		txt += 'cd $RUNTIME_AREA\n'
734	–	file_list=file_list[:-1]
735	–	txt += 'file_list="'+file_list+'"\n'
736	–	### OLI_DANIELE
737	–	txt += 'if [ $middleware == OSG ]; then\n'
738	–	txt += '    cd $RUNTIME_AREA\n'
739	–	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
740	–	txt += '    /bin/rm -rf $WORKING_DIR\n'
741	–	txt += '    if [ -d $WORKING_DIR ] ;then\n'
742	–	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
743	–	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
744	–	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
745	–	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
746	–	txt += '    fi\n'
747	–	txt += 'fi\n'
748	–	txt += '\n'
1218		return txt
1219
1220		def numberFile_(self, file, txt):
#	Line 756 | Line 1225 | class Cmssw(JobType):
1225		# take away last extension
1226		name = p[0]
1227		for x in p[1:-1]:
1228	<	name=name+"."+x
1228	>	name=name+"."+x
1229		# add "_txt"
1230		if len(p)>1:
1231	<	ext = p[len(p)-1]
1232	<	#result = name + '_' + str(txt) + "." + ext
764	<	result = name + '_' + txt + "." + ext
1231	>	ext = p[len(p)-1]
1232	>	result = name + '_' + txt + "." + ext
1233		else:
1234	<	#result = name + '_' + str(txt)
767	<	result = name + '_' + txt
1234	>	result = name + '_' + txt
1235
1236		return result
1237
1238	<	def getRequirements(self):
1238	>	def getRequirements(self, nj=[]):
1239		"""
1240		return job requirements to add to jdl files
1241		"""
1242		req = ''
1243	<	if common.analisys_common_info['sites']:
1244	<	if common.analisys_common_info['sw_version']:
1245	<	req='Member("VO-cms-' + \
1246	<	common.analisys_common_info['sw_version'] + \
1247	<	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1248	<	if len(common.analisys_common_info['sites'])>0:
1249	<	req = req + ' && ('
1250	<	for i in range(len(common.analisys_common_info['sites'])):
1251	<	req = req + 'other.GlueCEInfoHostName == "' \
1252	<	+ common.analisys_common_info['sites'][i] + '"'
1253	<	if ( i < (int(len(common.analisys_common_info['sites']) - 1)) ):
1254	<	req = req + ' || '
1255	<	req = req + ')'
789	<	#print "req = ", req
1243	>	if self.version:
1244	>	req='Member("VO-cms-' + \
1245	>	self.version + \
1246	>	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1247	>	## SL add requirement for OS version only if SL4
1248	>	#reSL4 = re.compile( r'slc4' )
1249	>	if self.executable_arch: # and reSL4.search(self.executable_arch):
1250	>	req+=' && Member("VO-cms-' + \
1251	>	self.executable_arch + \
1252	>	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1253	>
1254	>	req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1255	>
1256		return req
1257
1258		def configFilename(self):
#	Line 803 | Line 1269 | class Cmssw(JobType):
1269		txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
1270		txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1271		txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1272	+	txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
1273		txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1274	<	txt += '   elif [ -f $OSG_APP/cmssoft/cmsset_default.sh ] ;then\n'
1275	<	txt += '      # Use $OSG_APP/cmssoft/cmsset_default.sh to setup cms software\n'
1276	<	txt += '       source $OSG_APP/cmssoft/cmsset_default.sh '+self.version+'\n'
1274	>	txt += '   elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1275	>	txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1276	>	txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
1277	>	txt += '       source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1278		txt += '   else\n'
1279	<	txt += '       echo "SET_CMS_ENV 10020 ==> ERROR $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cmsset_default.sh file not found"\n'
1279	>	txt += '       echo "SET_CMS_ENV 10020 ==> ERROR $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1280		txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
1281		txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1282		txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1283	+	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1284	+	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1285	+	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1286		txt += '       exit 1\n'
1287		txt += '\n'
1288		txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1289		txt += '       cd $RUNTIME_AREA\n'
1290		txt += '       /bin/rm -rf $WORKING_DIR\n'
1291		txt += '       if [ -d $WORKING_DIR ] ;then\n'
1292	<	txt += '            echo "SET_CMS_ENV 10017 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cmsset_default.sh file not found"\n'
1293	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1294	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1295	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1292	>	txt += '           echo "SET_CMS_ENV 10017 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1293	>	txt += '           echo "JOB_EXIT_STATUS = 10017"\n'
1294	>	txt += '           echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1295	>	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1296	>	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1297	>	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1298	>	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1299		txt += '       fi\n'
1300		txt += '\n'
1301		txt += '       exit 1\n'
#	Line 845 | Line 1319 | class Cmssw(JobType):
1319		txt += '       echo "JOB_EXIT_STATUS = 10031" \n'
1320		txt += '       echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1321		txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1322	+	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1323	+	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1324	+	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1325		txt += '       exit 1\n'
1326		txt += '   else\n'
1327		txt += '       echo "Sourcing environment... "\n'
#	Line 853 | Line 1330 | class Cmssw(JobType):
1330		txt += '           echo "JOB_EXIT_STATUS = 10020"\n'
1331		txt += '           echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1332		txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1333	+	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1334	+	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1335	+	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1336		txt += '           exit 1\n'
1337		txt += '       fi\n'
1338		txt += '       echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
#	Line 863 | Line 1343 | class Cmssw(JobType):
1343		txt += '           echo "JOB_EXIT_STATUS = 10032"\n'
1344		txt += '           echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1345		txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1346	+	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1347	+	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1348	+	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1349		txt += '           exit 1\n'
1350		txt += '       fi\n'
1351		txt += '   fi\n'
1352		txt += '   \n'
870	–	txt += '   string=`cat /etc/redhat-release`\n'
871	–	txt += '   echo $string\n'
872	–	txt += '   if [[ $string = *alhalla* ]]; then\n'
873	–	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
874	–	txt += '   elif [[ $string = *Enterprise* ]] || [[ $string = *cientific* ]]; then\n'
875	–	txt += '       export SCRAM_ARCH=slc3_ia32_gcc323\n'
876	–	txt += '       echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
877	–	txt += '   else\n'
878	–	txt += '       echo "SET_CMS_ENV 10033 ==> ERROR OS unknown, LCG environment not initialized"\n'
879	–	txt += '       echo "JOB_EXIT_STATUS = 10033"\n'
880	–	txt += '       echo "JobExitCode=10033" | tee -a $RUNTIME_AREA/$repo\n'
881	–	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
882	–	txt += '       exit 1\n'
883	–	txt += '   fi\n'
1353		txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1354		txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1355		return txt
1356
1357	+	### FEDE FOR DBS OUTPUT PUBLICATION
1358	+	def modifyReport(self, nj):
1359	+	"""
1360	+	insert the part of the script that modifies the FrameworkJob Report
1361	+	"""
1362	+
1363	+	txt = ''
1364	+	try:
1365	+	publish_data = int(self.cfg_params['USER.publish_data'])
1366	+	except KeyError:
1367	+	publish_data = 0
1368	+	if (publish_data == 1):
1369	+	txt += 'echo "Modify Job Report" \n'
1370	+	#txt += 'chmod a+x $RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1371	+	################ FEDE FOR DBS2 #############################################
1372	+	txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1373	+	#############################################################################
1374	+	#try:
1375	+	#    publish_data = int(self.cfg_params['USER.publish_data'])
1376	+	#except KeyError:
1377	+	#    publish_data = 0
1378	+
1379	+	txt += 'if [ -z "$SE" ]; then\n'
1380	+	txt += '    SE="" \n'
1381	+	txt += 'fi \n'
1382	+	txt += 'if [ -z "$SE_PATH" ]; then\n'
1383	+	txt += '    SE_PATH="" \n'
1384	+	txt += 'fi \n'
1385	+	txt += 'echo "SE = $SE"\n'
1386	+	txt += 'echo "SE_PATH = $SE_PATH"\n'
1387	+
1388	+	#if (publish_data == 1):
1389	+	#processedDataset = self.cfg_params['USER.processed_datasetname']
1390	+	processedDataset = self.cfg_params['USER.publish_data_name']
1391	+	txt += 'ProcessedDataset='+processedDataset+'\n'
1392	+	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1393	+	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1394	+	#### FEDE: added slash in LFN ##############
1395	+	txt += '    FOR_LFN=/copy_problems/ \n'
1396	+	txt += 'else \n'
1397	+	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1398	+	#####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
1399	+	txt += '    FOR_LFN=/store$tmp \n'
1400	+	txt += 'fi \n'
1401	+	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1402	+	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1403	+	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1404	+	#txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1405	+	txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1406	+	txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1407	+	#txt += '$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1408	+
1409	+	txt += 'modifyReport_result=$?\n'
1410	+	txt += 'echo modifyReport_result = $modifyReport_result\n'
1411	+	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1412	+	txt += '    exit_status=1\n'
1413	+	txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
1414	+	txt += 'else\n'
1415	+	txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1416	+	txt += 'fi\n'
1417	+	else:
1418	+	txt += 'echo "no data publication required"\n'
1419	+	#txt += 'ProcessedDataset=no_data_to_publish \n'
1420	+	#### FEDE: added slash in LFN ##############
1421	+	#txt += 'FOR_LFN=/local/ \n'
1422	+	#txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1423	+	#txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1424	+	return txt
1425	+
1426	+	def cleanEnv(self):
1427	+	### OLI_DANIELE
1428	+	txt = ''
1429	+	txt += 'if [ $middleware == OSG ]; then\n'
1430	+	txt += '    cd $RUNTIME_AREA\n'
1431	+	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
1432	+	txt += '    /bin/rm -rf $WORKING_DIR\n'
1433	+	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1434	+	txt += '              echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1435	+	txt += '              echo "JOB_EXIT_STATUS = 60999"\n'
1436	+	txt += '              echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1437	+	txt += '              dumpStatus $RUNTIME_AREA/$repo\n'
1438	+	txt += '        rm -f $RUNTIME_AREA/$repo \n'
1439	+	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1440	+	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1441	+	txt += '    fi\n'
1442	+	txt += 'fi\n'
1443	+	txt += '\n'
1444	+	return txt
1445	+
1446		def setParam_(self, param, value):
1447		self._params[param] = value
1448
#	Line 896 | Line 1454 | class Cmssw(JobType):
1454
1455		def getTaskid(self):
1456		return self._taskId
1457	+
1458	+	def uniquelist(self, old):
1459	+	"""
1460	+	remove duplicates from a list
1461	+	"""
1462	+	nd={}
1463	+	for e in old:
1464	+	nd[e]=0
1465	+	return nd.keys()
1466	+
1467	+
1468	+	def checkOut(self, limit):
1469	+	"""
1470	+	check the dimension of the output files
1471	+	"""
1472	+	txt = 'echo "*****************************************"\n'
1473	+	txt += 'echo "** Starting output sandbox limit check **"\n'
1474	+	txt += 'echo "*****************************************"\n'
1475	+	allOutFiles = ""
1476	+	listOutFiles = []
1477	+	for fileOut in (self.output_file+self.output_file_sandbox):
1478	+	if fileOut.find('crab_fjr') == -1:
1479	+	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob')
1480	+	listOutFiles.append(self.numberFile_(fileOut, '$NJob'))
1481	+	txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1482	+	txt += 'ls -gGhrta;\n'
1483	+	txt += 'sum=0;\n'
1484	+	txt += 'for file in '+str(allOutFiles)+' ; do\n'
1485	+	txt += '    if [ -e $file ]; then\n'
1486	+	txt += '        tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n'
1487	+	txt += '        sum=`expr $sum + $tt`\n'
1488	+	txt += '    else\n'
1489	+	txt += '        echo "WARNING: output file $file not found!"\n'
1490	+	txt += '    fi\n'
1491	+	txt += 'done\n'
1492	+	txt += 'echo "Total Output dimension: $sum";\n'
1493	+	txt += 'limit='+str(limit)+';\n'
1494	+	txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1495	+	txt += 'if [ $limit -lt $sum ]; then\n'
1496	+	txt += '    echo "WARNING: output files have to big size - something will be lost;"\n'
1497	+	txt += '    echo "         checking the output file sizes..."\n'
1498	+	"""
1499	+	txt += '    dim=0;\n'
1500	+	txt += '    exclude=0;\n'
1501	+	txt += '    for files in '+str(allOutFiles)+' ; do\n'
1502	+	txt += '        sumTemp=0;\n'
1503	+	txt += '        for file2 in '+str(allOutFiles)+' ; do\n'
1504	+	txt += '            if [ $file != $file2 ]; then\n'
1505	+	txt += '                tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1506	+	txt += '                sumTemp=`expr $sumTemp + $tt`;\n'
1507	+	txt += '            fi\n'
1508	+	txt += '        done\n'
1509	+	txt += '        if [ $sumTemp -lt $limit ]; then\n'
1510	+	txt += '            if [ $dim -lt $sumTemp ]; then\n'
1511	+	txt += '                dim=$sumTemp;\n'
1512	+	txt += '                exclude=$file;\n'
1513	+	txt += '            fi\n'
1514	+	txt += '        fi\n'
1515	+	txt += '    done\n'
1516	+	txt += '    echo "Dimension calculated: $dim"; echo "File to exclude: $exclude";\n'
1517	+	"""
1518	+	txt += '    tot=0;\n'
1519	+	txt += '    for file2 in '+str(allOutFiles)+' ; do\n'
1520	+	txt += '        tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1521	+	txt += '        tot=`expr $tot + $tt`;\n'
1522	+	txt += '        if [ $limit -lt $tot ]; then\n'
1523	+	txt += '            tot=`expr $tot - $tt`;\n'
1524	+	txt += '            fileLast=$file;\n'
1525	+	txt += '            break;\n'
1526	+	txt += '        fi\n'
1527	+	txt += '    done\n'
1528	+	txt += '    echo "Dimension calculated: $tot"; echo "First file to exclude: $file";\n'
1529	+	txt += '    flag=0;\n'
1530	+	txt += '    for filess in '+str(allOutFiles)+' ; do\n'
1531	+	txt += '        if [ $fileLast = $filess ]; then\n'
1532	+	txt += '            flag=1;\n'
1533	+	txt += '        fi\n'
1534	+	txt += '        if [ $flag -eq 1 ]; then\n'
1535	+	txt += '            rm -f $filess;\n'
1536	+	txt += '        fi\n'
1537	+	txt += '    done\n'
1538	+	txt += '    ls -agGhrt;\n'
1539	+	txt += '    echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1540	+	txt += '    echo "JOB_EXIT_STATUS = 70000";\n'
1541	+	txt += '    exit_status=70000;\n'
1542	+	txt += 'else'
1543	+	txt += '    echo "Total Output dimension $sum is fine.";\n'
1544	+	txt += 'fi\n'
1545	+	txt += 'echo "*****************************************"\n'
1546	+	txt += 'echo "*** Ending output sandbox limit check ***"\n'
1547	+	txt += 'echo "*****************************************"\n'
1548	+	return txt

Diff Legend

–	Removed lines
+	Added lines
<	Changed lines
>	Changed lines